Abstract: Converter lining is regarded as a critical structure for guaranteeing safe and stable operation of steelmaking processes. Its service state is directly reflected in converter campaign life, refractory consumption and production safety. During cyclic charging, blowing and tapping operations, the lining is subjected to coupled influences of multiple factors including attack by high-temperature molten slag and liquid steel, mechanical impact and scouring, thermal cycling and structural constraints. Abnormal erosion, cracking and spalling are easily induced in local zones, and burn-through and steel breakout accidents can be triggered under severe conditions. Systematic reviews are conducted on research progress concerning erosion mechanisms, erosion state monitoring and burn-through early warning technologies for converter linings. First, major damage mechanisms of converter linings are summarized from the perspectives of chemical erosion and slag penetration, mechanical impact and abrasion, thermal shock damage and structural stress. Regional and staged characteristics of erosion behaviours in typical zones are analysed, including the slag line zone, trunnion zone, tapping spout, furnace bottom and charging side. Second, influences of multiple factors on lining erosion are elaborated, such as smelting temperature and duration, MgO content, total iron content and basicity of slag, charging schedule and blowing operations. On this basis, fundamental principles, technical characteristics and engineering application status of lining condition monitoring and burn-through early warning approaches are emphatically reviewed, covering infrared thermography, laser thickness measurement, electrical detection, acoustic detection, neural network prediction and multi-source information fusion. Findings from existing investigations indicate that infrared thermography is adopted for continuous on-line monitoring of hot spots on the furnace shell. Laser thickness measurement is employed to quantitatively characterise residual lining thickness and local weak zones. Certain application potentials are demonstrated by electrical detection, acoustic detection and multi-source information fusion methods in continuous perception and improvement of early warning reliability. Nevertheless, popularisation and application of relevant technologies are still restricted by problems such as complicated on-site environments, temporally and spatially discontinuous monitoring data, insufficient cross-campaign adaptability of models and the absence of unified criteria for early warning thresholds. In future, deep integration between lining erosion mechanisms and data-driven models should be strengthened. Promotions should be realised for coordination of multi-source monitoring information, joint characterisation of residual thickness and erosion rate, and construction of a zoned and graded dynamic early warning system. Supports can thereby be provided for predictive maintenance and safe long-service operation of converter linings.